
*****************************************************************************
*
* Readme file for
*
* RESIGNATION AS PROMOTION? EXECUTIVE TURNOVER AND EARLY DEPARTURES IN THE ARGENTINE CONGRESS, 1983-2017
*
* By Adrián Lucardi, Juan Pablo Micozzi and Natán Skigin
*
* Legislative Studies Quarterly
*
* December 2021
*
*****************************************************************************


*****************************************************************************
* NOTE
*****************************************************************************

- All the analyses were carried out using R version 4.0.3 (2020-10-10) -- "Bunny-Wunnies Freak Out"
- The script was run on a Mac platform
- WARNING: both the Callaway and Sant'Anna estimates in the main (monthly) analysis and the daily analysis take A LOT of time --the actual value depends on your computer, but in a MacBook Air (Retina, 13-inch, 2018) with a 1.6 GHz Dual-Core Intel Core i5 processor and 8GB RAM it took a couple even days. This is because (a) Callaway and Sant'Anna's estimator works slowly for unbalanced panels; and (b) once cleaned, the daily dataset had 2.7 million legislator-day observations, with fewer than 300 resignations.
- To replicate the results, you need to install and call the following packages and dependencies in R:
install.packages ("pacman")
install.packages ("colourvalues")
install.packages ("did")
install.packages ("dplyr")
install.packages ("DRDID")
install.packages ("ggplot2")
install.packages ("ggrepel")
install.packages ("jtools")
install.packages ("lfe")
install.packages ("readxl")
install.packages ("survival")
install.packages ("stringr")
install.packages ("tidyr")
install.packages ("tidyverse")
install.packages ("wesanderson")
install.packages ("xtable")




*****************************************************************************
* REPLICATION FILES
*****************************************************************************

1. "Resignation_LSQ_01_Monthly.R": script used to reproduce the bulk of the results, tables and figures reported in the paper. The only exceptions are Tables A9 and A10 in the online appendix. This also produces the LaTeX code for the corresponding tables, though you have to copy and paste it manually.

2. "Resignation_LSQ_02_Daily.R": script used to reproduce the models reported in Tables A9 and A10 in the online appendix --i.e., the daily data models. This also produces the LaTeX code for all the corresponding tables, though you have to copy and paste it manually.

3. "Legislator-mandate full 1983-2017.RData": legislator-mandate dataset. List of Argentine deputies elected between 1983 and 2013, who thus ended their mandate between 1985 and 2017. Each row is a national deputy during a given mandate; the same individual may thus appear multiple times iff (s)he was elected to the Chamber more than once. Used to replicate:
- Table 1
- Figure 2a
- Table A1a
- Figure A2
- Figure A3

4. "Legislator-month full 1983-2017.RData": legislator-month dataset. Expanded version of the legislator-mandate dataset so that every row is a different legislator-month combination. Used to replicate:
- Figure 2b
- Figure 4
- Figure 5
- Figure 6
- Table A1b
- Figure A4
- Tables A2 through A8
- Figures A6 through A14

5. "Legislator-day full 1983-2017.RData": legislator-day dataset. Expanded version of the legislator-mandate dataset so that every row is a different legislator-day combination. Used to replicate:
- Figure 3
- Figure A5
- Table A9
- Tabla A10


6. "Careers (all) 1983-2017.xlsx": list of individuals who occupied the following offices between 1983 and 2017: (vice-)president; (deputy) governor; mayor; national minister; national senator; national deputy; national minister; member of the 1994 constituent assembly; and member of the Mercosur Parliament. Every row is an individual during a given mandate; reelected individuals thus appear multiple times, one for each mandate. Source: Juan Pablo Micozzi and Adrián Lucardi, "How Valuable Is a Legislative Seat? Long-Term Incumbency Effects in the Argentine Chamber of Deputies", Political Science Research and Methods, 9(2), 2021, pp. 414-29. Used to replicate:
- Figure A1
